Nemesys - Neural emotion eliciting system

This paper describes the development of a new model of agent emotion elicitation called Nemesys. It enhances interfaces with emotional and social information. Nemesys is based on an artificial neural network and is able to learn six basic emotional states. The elicitation of emotions is based on models drawn from the state of the art in modeling emotions in the field of psychology. Further the described framework includes the Five-Factor Model of Personality to represent different agent personalities. Nemesys (called after Nemesis, the Greek goddess of righteous anger) is designed to perform in various types of interfaces. The usage of Nemesys is presented with an application scenario employing a commercial 3D game-engine. Additionally a critical review of the current elicitation behavior of Nemesys is presented and discussed.